I think we should be more ant-like!

I recently moved to New York City, and since I've been here, I've watched 3 documentaries about ants. I cannot lie, the city reminds me of an ant hill. Ever buzzing, large, distributed, ordered but chaotic. This is actually an appealing thing for me. I love ants. They are underground fungus gardeners and aphid sheperds. The only bug to raise another bug as cattle. But most of all, I am impressed by their trust in one another. To such a high degree do they trust, that the term 'hive-mind' is often applied to them -- they act as if they share a collective conciousness. Rather than worrying about 'becoming ants' in the Orwelian sense, I wonder how individuals can embrace each others individuality so fully that they trust them to build, implement, and live without strict regulation to impede the process. Indeed, I must admit that while I complain about the overregulation and deep bureacracy in place to control how the poor spend their welfare checks, I am quite pleased that we have strict regulations on the oil industry -- and I wish they were stricter! I don't trust oil tycoons, it is my personal bias. But if somehow I could trust that I, like the poor, and the oil tycoons were all just part of a great distributed system relying on mutual trust, maybe, we'd all be better off.

Enormous leaf-cutter anthill excavation

Welcome to the ant hill

Ants are cool, for all sorts of reasons. Perhaps one of the most interesting things about them, is that despite the enormous size of their colonies -- some super colonies have 300 million ant or more -- they do not require a centralized controlling body to govern their society. It is deceiving that we refer to a certain ant as the Queen. It is true she is important, but only because she births the entire colony -- she defers all decision making to her children.

If an ant hill was a trivial thing to run, perhaps this would not excite me so much, but believe me, it is not. For example, in a single leaf cutter ant hill, a whole host of roles exist for worker ants. Cutter Ants slice through leaves with their strong mandibles and let them fall to the ground for Transporter Ants to carry back to the hill (marching 2 by 2 hurrah! hurrah!). Gardeners, take the leaves into the depths of the colony, bringing them to specialized gardening rooms where the leaves are left to decompose and become fertilizer for a special fungus that the Gardener Ants grow as food. Since leaf decomposition and fungus growth produces carbon dioxide which is poisonous to oxygen breathing ants, the colony must be built so that CO2 leaves the hill, and oxygen replaces it. Since there is no 'Architect Ant' specialty, all ants must assist in tunnel design and digging whenever necessary.

Furthermore, the number of ants needed to fill each role changes based on circumstance. Leaf cutters have another type of subteranean worker called a Soldier Ant. Soldiers each patrol a small area of the colony, walking back and forth like a dilligent mall cop. If a disturbance happens anywhere in the colony though, the Soldiers are quick on the scene, and very effective (they've been documented to have successfully protected colonies from bears). The number of Soldiers necessary in a colony is not static of course -- it waxes and wanes depending on current threats. Amazingly, with no centralized authority, ants generally have the right number of soldiers, not too many and not to few. The Nursery Ants, who spend their days rearing their larval sisters, feed the larvae according to their future specialty, and adjust their numbers based on the current needs of the colony. Ants fed in one way will become large Soldiers, others efficent and speedy Transporters. Still another might be fed a diet that will make her into her mother's successor.

Ruling with no King

Over Christmas, my Aunt Christine gifted me with a collaborative game called Hanabi. The goal of this card game is to 'create a fireworks display' by placing cards on the table in the correct order. Each player can see her teammates cards, but not her own. The card deck consists of five different colors of cards, numbered 1–5 in each color. For each color, the players try to place a row in the correct order from 1–5. The cards have different numbers based on value, but there are some duplicates. It takes 25 cards to win the game, and the deck is 50, which means up to 25 can be discarded. There are a few things working against you though.

  1. Information sharing is limited. Sharing information with a teammate about her cards is the key to getting her to play the right ones in the future, but it's not a free for all. Information sharing costs a Time Token. Since Time Tokens can only be bought by discarding cards, and since a limited number of cards can be discarded, there is only so much you can say. Also, the pieces of information that are shareable follow a strict set of guidlines.

  2. Mistakes will cost you. Since the information highway in this game has the bandwith of a narrow mountain pass, it makes some sense to take some chances. After all, out of 50 cards, only 25 are needed to win the game right? Not quite. 3 mistakes and you're done, that's all you get. Also, since there is only 5 5's in the whole game if you throw one of these away by accident, your ruined fireworks blow you up immedietly. Tough luck Uncle Sam.

It's a trust game. The key, is to give your partner large pieces of information rather than specific and obvious ones. This takes full advantage of our limited information sharing bandwith. It's hard to do though, for those of us who like being in control, because it leaves the decision making completely up to our teammate.

For example, I might notice my partner has a yellow 2 that is actionable -- there is a yellow 1 on the board. I might be tempted to tell him 'YOU HAVE A 2' loudly, in hopes that he will confidently play it assuming that I have not told him such specific information without purpose. This will probably work for this turn, but in sharing with your teammate such limited information about his hand, you have made it so he is in the dark about his hand once again after this turn! Had you only had patience you could have maximized instead of specified.

In Hanabi, information sharing means revealing all of the cards of a specific color or number. So you can say, 'THESE 3 CARDS ARE YELLOW'. Next turn, I might tell him which of his cards match a number, and save my highly specific information for times of truly high impact, ie 'THIS CARD IS A 5'. By maximizing information tidbits we create a highly informed teammate, and can trust them to make the best decision because of the information we've given them. By saying 'YOU HAVE A 2' we are attempting to be the King. By maximizing information flow, we are saying, I trust you to dig the tunnel.

Relinquishing control means radical trust, and hopefully, radical efficency

Trust is efficent. If you've seen 'Split or Steal', read about Game Theory, or seen A Beautiful Mind, you know this. Recently, I stumbled upon a videogame version of some game theory problems which is beautifully illustrated and very thought provoking.


It's also risky. Trust means you are not in control all the time, and it means you can get screwed. Naivete is not what I preach, but I do enjoy exploring what it would mean to employ trust as a convention in more places. Trust is already the convention in more places than you might think about -- from red lights and stop signs, to wall mounted shelves sized according to the distance between studs specified in local building codes, to software that relies on other software and trusts it's semantic versioning schemes and API documentation . Imagine the ineffient society that must create a temporary impassible barrier in order to avoid accidents at even the quietest intersections. Imagine the time and money that could be saved if we trusted each other in more places.

Start the movement by starting

Last evening, on the subway, a man boarded and asked for change. As someone new to NYC, I imagined this was an incredibly common occurrence and expected he'd be ignored. I'm sure that on other trains in the city this is often the case. That night though, a great number of different people gave him change, not a single one gave him paper money, but so many handed him something and each one looked in his eye and wished him luck. I only had a 20 dollar bill in my pocket and was unwilling to part with it that evening, but the next day, I filled my pocket with some change, hoping to be part of the small trust based system that had doubtlessly amassed enough change to make an out-of-luck man's day a little bit better. While each one could have thought his 2 quarters unworthy and so not bothered to give them, the collective effort had impact, on the man, and also, on my future behavior.

The first person who gave change to this man had the most trust. She was trusting that someone else on the train would give money too and make her money less worthless and more impactful. Had she not started by eagerly giving some change, the others might not have given any money. By being the first to give a few quarters on a train, she gave a man enough for a meal.

The lesson here is two-fold. Start the movement by starting, and be part of the movement when it's happening.

Basically, when JR Smith does the no-look lob, you better Alley Oop, that shit like you're LeBron James.


Prev <> Next


(c) 2018 maxbuckdeveloper@gmail.com